To innovate, we need to try things out; to experiment.

But how do we select the right experiment? How do we choose?

Whenever we conduct an experiment, we want the maximum information at the minimum cost. Experiments should be as simple and cheap as possible, but sophisticated enough to answer your questions.

In vivo, in vitro or in silico experiments

I find it useful to think about business experiments the same way they do in the biological sciences. Experiments can be in vivo, in vitro, or in silico.

  • In vivo – in life – experiments carried out live in the final target application;
  • In vitro – in glass – experiments carried out in the lab, simplified and in a controlled environment;
  • In silico – in a computer – experiments carried out using simulation and modelling.

A business experiment, like a biological experiment, can fall into any of these three categories. What drives your choice is:

  • The quality of the information you can get; how useful it is to your decision-making;
  • How practical the experiment is;
  • How expensive the experiment is, in time, resources and money.

The value for money sweet spot can be anywhere between full digital simulation and live in-use testing.

“Somewhere on the line between computer simulation and in-use trials you will find the best value for money.”

Real-life and in-use testing is best if it is possible

Weihenstephan bottling plant

If you can manage it, a real-life test is the best. You will gain much more information, and more important the trial will be much more convincing for your target customers. Testing a new surface coating for a paper mill, or an adhesive for a high-speed bottling machine, really needs a production line. You will need a good relationship with your customer, and some good evidence, before they will let you near the real application. You could cost them a lot of money and lost production if it goes wrong, and they will have to trust you.

Live production line testing represents a risk for your customers, but there are plenty of innovations you can put directly into the hands of the target users to see what happens. If the cost to you and the risks to the customer are low, you can test market, or give away free samples in exchange for customer feedback.

Experimenting under simplified and controlled conditions

Your product may not be ready for real-life testing, or there may be regulatory barriers that prevent test marketing. In that case, you can bring the experiment into a controlled environment with careful supervision. A focus group or consumer panel can evaluate your concept or prototype.

Alternatively, you may want to observe how the target customers use your innovation. What works and what doesn’t work? The unexpected ways that users interact with the product. Even when done in the field, such experiments are more controlled and simplified than live testing.

And in most innovations, there will be a lot of testing of components, sub-systems and prototypes before a product is ready to show to the end-user. All under controlled conditions.

Computer simulation lets you experiment with complex systems

There are situations where any kind of live testing is too dangerous or impractical. We need to know if a nuclear power station will fail-safe under all conditions, but you cannot carry out the experiment in case something does go wrong.

A city-wide change to the transport system may bring real benefits, but it is not practical to live-test it before implementation. You can test some sub-systems in the lab, but that won’t tell you about all the complex interactions in the real city.

Computer modelling and simulation come into their own when physical experiments are not possible.

Simulation is also useful when you want more detailed information than can be easily gained from a physical experiment. A good example is the use of computational fluid dynamics to understand vehicle drag and energy efficiency.

With rapid developments in computer power and software, we can simulate more and more innovation problems, but it is not the answer to all questions. It works well for many engineering applications and biological systems, but it is still time-consuming, expensive and difficult to simulate a complex system like a city.

Choose experiments wisely

So when developing your business experiments, think about where you are on your innovation journey, and what is the key question you need to answer to make progress? Somewhere on the line between computer simulation and in-use trials you will find the best value for money. The most useful information per dollar spent.

How to Pick the Right Business Experiment
Tagged on:         

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.